Littlewood's Law and the Global Media

post by gwern · 2019-01-12T17:46:09.753Z · score: 37 (8 votes) · LW · GW · 3 comments

This is a link post for https://www.gwern.net/Littlewood

3 comments

Comments sorted by top scores.

comment by Benito · 2019-01-13T10:37:02.783Z · score: 7 (3 votes) · LW · GW

Wow. Reading the story of Air France 447 was intense (and horrifying). It's remarkable how many little things went wrong that didn't need to. It felt weirdly like reality was conspiring to make the plane crash ("Just before the critical moment, let's send the most experienced pilot off for a nap" and "During the critical moment, let's make sure the pilot in command mentions all the information except the critical piece that could save them" etc), even though I recognise that it's all a selection effect on picking the one crash out of 10s of billions of very safe flights. I'll definitely keep Littlewood's Law in mind when discussing the news in the future.

The coping mechanisms for the news feel optimistic. Things like 'screen size' or 'typography' in proportion to base rate seems unlikely to produce a good product that people can use. I think to solve news you probably need a high effort process that gathers lots of information and doesn't have to compete quickly for attention and can sort through it and make a decision about what to publish, rather than trusting all the consumers of news to do the statistical data gathering themselves, each individually largely getting it right. (A while back I saw this new org building itself around a subscription based system rather than an ads based system, and after ~20 mins of exploring their site I paid to be a member, though I don't expect to read it.)

I've put a bunch of work myself into basically not updating when reading a news article on basically anything except the editorial process of the news org. It feels to me related to when a character reads a news article in a movie, or perhaps when I read fanfiction and learn new info about a character that isn't canon. I'm definitely building a web of implications and associations, but it's very disconnected from my model of the real world. And obviously more than that I've put a bunch of work into not being exposed to the news in general.

Unrelated: I hadn't seen this before, and it is a fun resource to explore.

comment by quanticle · 2019-01-13T09:03:01.410Z · score: 5 (2 votes) · LW · GW

I was able to use this post when discussing the news with a family member of mine. The example of a one-in-a-million event occurring 8 times a month (plus increasing global interconnection ensuring that we hear about these events every time they do occur) was especially helpful in helping debiasing someone who had read too much of the news.

comment by Dagon · 2019-01-14T21:00:39.985Z · score: 3 (2 votes) · LW · GW
one-in-a-million event occurring 8 times a month

Off by a thousand. One-in-a-million happens 8 times in Manhattan and a thousand times in China. One-in-a-BILLION is 8 times worldwide.